Simultaneous masking is a property of the human auditory system where some sounds simply vanish in the presence of other sounds with certain characteristics (so called maskers).
In the model described in the implemented coder, a short-term frequency representation of the audio is used to help estimate the masking function. To get a frequency representation, a 512-sample FFT with a sine window is performed on the audio every 256 samples to form . Every frequency bin is mapped to a corresponding critical band (a real number), using (8).